In this study, we demonstrated that adequate quality spirometry can be obtained in three-quarters of hospitalized patients with a physician diagnosis of asthma or COPD exacerbation. Spirometry confirmed obstructive lung disease in 78% of participants with adequate quality tests; in other words, about 1 in 5 participants with a physician diagnosis of asthma or COPD exacerbation did not meet the diagnostic criteria by spirometry. Overdiagnosis was about four times more likely in obese than in non-obese patients.
Our finding demonstrates that adequate quality spirometry can be obtained early in the hospital course in most patients with asthma or COPD exacerbations. This finding expands previous work by Rea and colleagues
, who showed that patients hospitalized with COPD exacerbations were able to perform spirometry on the day of discharge. None of the participant characteristics we examined were different between those who were able to produce adequate vs. not adequate quality spirometry, except for a trend between the number of days between hospital admission and spirometry testing. We found a trend suggesting that a shorter number of days between hospital admission and spirometry testing may be associated with the inability to produce adequate spirometry. We suspect that the length of hospital stay before spirometry testing is linked to the acuity of illness (i.e. sicker patients were unable to perform spirometry). Therefore, an important limitation to our study is that patients who were too ill to provide informed consent and those who were admitted to the intensive care unit were excluded; such patients may be even more likely to not perform adequate quality spirometry tests.
Recent studies of spirometry in the primary care setting have shown that up to 50% of patients with a physician diagnosis of COPD did not meet the criteria for the diagnosis by spirometry
[21–23]. Similarly, a recent study showed that more than 30% of patients with a physician diagnosis of stable asthma did not meet the criteria for the diagnosis when tested with a combination of spirometry and methacholine challenge test
. To our knowledge, the present report is the first study to evaluate the frequency of misclassification among hospitalized patients with a physician diagnosis of asthma exacerbation.
Additionally, we found that the lack of evidence for obstructive lung disease was four times more common in obese vs. non-obese participants. Our findings in hospitalized patients are consistent with results observed in a study of outpatients in a primary care setting that identified higher rates of misclassification for COPD in overweight or obese patients
. Similarly, previous studies have suggested that overweight or obese patients may be over-diagnosed as having asthma. For example, one study reported that nearly 1 in 3 overweight or obese subjects diagnosed with asthma do not actually have airway hyperresponsiveness
. Other studies, including some in obese patients, found that medical history and physical examination findings may not be sufficiently reliable to diagnose obstructive lung disease
[26, 27]. Other data suggest that vocal cord dysfunction may mimic an asthma exacerbation
, which may help explain why some spirometry tracings had evidence of variable extrathoracic airflow obstruction. In our study, we found that some patients hospitalized with a diagnosis of asthma or COPD exacerbation presented spirometry tracings suggestive of restrictive (not obstructive) lung disease. Thus, together with previously published evidence, our findings suggest that a range of conditions may be contributing to respiratory symptoms diagnosed as asthma or COPD exacerbations.
This multicenter study had multiple strengths. First, spirometry was performed on average within 1 day of hospitalization, decreasing the likelihood that participants were tested after the resolution of the exacerbation. Second, we prospectively identified patients with a physician diagnosis of asthma or COPD exacerbation and in whom the treating physician did not suspect other respiratory conditions, rather than relying on chart abstraction or billing codes. Lastly, we employed standardized spirometry procedures using ERS/ATS guidelines, minimizing any variations in the procedure or interpretation.
There were also several potential limitations to our study. We relied on spirometry to diagnose obstructive lung disease and may have missed mild air trapping or increased peripheral airway resistance that would require the use of special diagnostic tests, such as body plethysmography or impulse oscillometry. However, ERS/ATS guidelines for the diagnosis of asthma or COPD are based on spirometry and do not require the use of such tests. Also, as participants were enrolled with respiratory symptoms requiring hospitalization, the likelihood of milder forms of airway disease requiring more sensitive tests would be very low. We relied on self-reported height and weight to determine the predicted spirometry results, which may not be as reliable as objective measurement of height and weight. We also enrolled a modest number of participants (n = 113) in four academic medical centers and we do not know if these results would be generalizable to other healthcare institutions (e.g., community hospitals). However, our results highlight the need for larger, multi-center studies to further evaluate the overdiagnosis of asthma and COPD exacerbations in hospitalized patients. Last, we only had limited data about patient characteristics. Additional information on the exacerbation severity and cognitive ability of the participants could have helped identify other characteristics associated with the inability to obtain adequate quality tests. We did not collect data regarding the comorbid conditions in patients without evidence of obstructive lung disease by spirometry (e.g., heart failure); this information may have offered clues into alternative diagnoses. Further, it is possible that the high proportion of obese patients without evidence of obstructive lung disease is due to residual confounding factors, such as respiratory muscle weakness, that were not measured in our study. Findings in this report can help to inform the design of larger multi-center, longitudinal studies that include community hospitals to assess differences in accuracy across institutions and within subgroups of patients.